The goals of the proposed study are to conduct an effectiveness trial of the keepin'it REAL middle school substance use prevention curriculum among a new target audience in rural Pennsylvania and Ohio, describe how teachers adapt the curriculum when they present it, and develop, implement, and evaluate a Pennsylvania/Ohio-version of the curriculum to test whether an evidence-based universal curriculum can be improved by adapting it to local cultures. keeping'it REAL is recognized as a model program by SAMHSA's National Registry of Effective programs and is one of the few that are multicultural. The study will evaluate the effectiveness of the original curriculum, grounded in the cultures of the southwest and compare that to a new version, reground in the rural culture of Pennsylvania and Ohio, while studying how teachers adapt both versions. This proposal responds to NIDA PA-05-118, Drug Abuse Prevention Intervention Research that calls for investigations addressing, 1) the development of novel drug abuse prevention approaches;2) the efficacy and effectiveness of newly developed and/or modified prevention programs;3) the processes associated with the selection, adoption, adaptation, implementation, sustainability, and financing of empirically validated interventions. This proposal addresses all three points. A randomized control trial will be conducted in middle schools to accomplish these goals. First, formative research will be conducted to develop a rural Pennsylvania/Ohio-version of the curriculum. Second, 42 rural schools will be randomly assigned to one of three conditions: teacher adaptation in which the original keepin'it REAL curriculum is implemented;researcher adaptation in which a new Pennsylvania-version of the curriculum is implemented, and a control group. We hypothesized the participation in either form of the curriculum will reduce drug use and that the researcher adaptation will produce better outcomes and less teacher adaptation than the teacher adaptation. A pretest will be administered followed by posttests in 7-9th grades. Adaptation and fidelity will be measured in 3 ways: teachers completing a Program Quality and Adaptation online measure after each lesson, videotaped lessons, and attendance. The major hypothesis tests will be conducted using variants of the general linear model, taking into account the multilevel structure of the data (e.g., multilevel multiple regression), test of a mediation model, and growth modeling. The development of effective school-based substance use prevention programs has given rise to a recent focus on implementation issues. This study will help us understand how to implement programs beyond their original target audience as well as how to guide teachers when they adapt curriculum to their classes.